Fig. 3

SMPP prediction model ROC curve. (a) The ROC curve shows an AUC of 0.985, with a 95% CI of 0.972–0.997, a 0.222 cut-off value, and sensitivity and specificity of 0.926 and 0.941, respectively. (b) The Hosmer–Lemeshow goodness-of-fit test results show that the nomogram model predicted SMPP risk with a good consistency (χ2 = 2.127, P = 0.977)

* AUC: Area under the curve, SMPP: Severe mycoplasma pneumoniae pneumonia